feature: allow wasm-bindgen to work on emscripten targets (INCOMPLETE/RFC)

[durin42]

We've got this wired up on our internal build system and wanted to try and get feedback before we get further. With our internal build system we've got this working end-to-end, and we're happy to do work here to make sure the right things happen with cargo including for cargo test. The question right now is: are y'all open to this landing if we're doing the work, or is this completely out of scope for what you want in the project?

(I've seen some other conversations about this in old bugs, but don't know the current direction of the project.)

Thanks!

[durin42]

@daxpedda could you let me know if this is a viable direction, or something we wouldn't be able to land? Like I said, we're happy to do the work, but don't want to spend the time if emscripten support will be rejected.

[daxpedda]

Similar things have been discussed about WASI here as well, see #3454.

In general I'm not against Emscripten support, but I'm not particularly in favor of it either. My understanding is that it should be possible to stick with wasm32-unknown-unknown and consume Emscripten dependencies (e.g. C/C++). Anything that helps that along I'm happy to improve on. But this isn't a reason to block support, just explains the lack of driving force behind it.

On the other hand I have so little knowledge about Emscripten that I honestly would be unable to form a qualified opinion here and I'm not interested in diving into it. So this needs a different maintainer!